Google issues AI playbook for engineers to boost productivity as Microsoft mulls grading employees on AI use

India Today

13 hours ago

Google has taken a major step to formalise how its software engineers use artificial intelligence at work. In an internal email sent earlier this week, the company shared detailed guidelines on how employees should adopt AI tools to boost productivity. The move comes as Google CEO Sundar Pichai recently revealed that over 30 per cent of all code written at the company is now AI-generated — a sharp rise from 25 per cent just a few months ago. Last month, the Google CEO also said that AI tools have helped its engineers be 10 per cent more productive.advertisementThe guidance, described by 9to5Google as the most formal communication on this topic so far, outlines best practices for using AI in day-to-day coding. Created by engineers at Google, the playbook reflects what has worked well internally and aims to help teams across the organisation work more efficiently.'We want every engineer to use AI to move faster and build better,' said a Google spokesperson. 'These recommendations and best practices, developed by a group of software engineers at Google, are designed to help all SWEs [software engineers] across the company get the most out of AI in their day-to-day work.'
Google's internal tools like 'Goose', which are trained on years of technical data, are already assisting engineers in writing code, fixing bugs, and even handling routine development tasks. Externally, Google has rolled out tools such as Gemini Code Assist, Gemini CLI, and Gemini in Android Studio. But internally, the company is now doubling down on the expectation that everyone must embrace these tools.advertisementThe new guidance not only focuses on AI-powered coding but also helps engineers understand the limitations of today's models. Managers and tech leads are being encouraged to integrate AI practices into their team workflows. While code is increasingly AI-generated, human oversight remains essential. The guidance stresses maintaining rigorous code review, testing, and security standards even with AI in the loop.Pichai has said AI tools have already improved engineering productivity by around 10 per cent. In a podcast with researcher Lex Fridman, he explained how Google measures time saved through AI and reinvests it into more valuable work. He called AI an 'accelerator'. He says it is a tool to free up engineers from repetitive tasks so they can focus on creative problem-solving.Still, the broader industry is wrestling with how to scale AI use internally. Microsoft, for instance, is reportedly planning to evaluate employees based on how effectively they use AI tools like GitHub Copilot. According to Business Insider, Microsoft has instructed team managers to include AI usage in performance reviews. Julia Liuson, President of Microsoft's Developer Division, wrote in an internal email: 'AI is now a fundamental part of how we work It's no longer optional — it's core to every role and every level.'advertisementThis move follows concerns inside Microsoft that internal adoption of Copilot has not met expectations. To change that, the company wants AI use to become as routine as collaboration or communication, especially among teams building AI products.Even Meta's Mark Zuckerberg has said that AI could soon handle 100 per cent of his company's developer work, highlighting how central AI has become to software development across the tech world. Meta on Monday also announced the Meta Superintelligence Labs, which is aimed at developing artificial general intelligence.Still, Pichai has tried to strike a balance. He says Google will continue hiring engineers and sees AI as a tool that complements, not replaces, human talent. But he also acknowledged the wider debate around job displacement. Citing remarks from Anthropic CEO Dario Amodei, who predicted that AI could replace half of entry-level white-collar jobs within five years, Pichai said it's important to have open discussions about AI's impact on employment, especially as it begins to automate more white-collar tasks.- Ends